This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.
When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:
url <- 'https://docs.google.com/spreadsheets/d/1DE08OpdFC2eH8VwAV-xLmsFByGh2VvL0DRDjPLhyo9U/edit?usp=sharing'
dat <- gsheet2tbl(url)
dat_clean <- dat %>%
filter(!is.na(author) & !author == "test" & !author == 0)
dat_clean$cohort <- paste(dat_clean$author, dat_clean$year, dat_clean$cohort_id, sep = "_")
dat_pintens <- dat_clean %>%
filter(measurements_pain_intensity == "Yes")
ts_m <- select (dat_pintens, cohort, author, year, cohort_id, pintens_pre_m, pintens_post_m, pintens_fu1_m,
pintens_fu2_m, pintens_fu3_m, measurements_pain_intensity, pintens_name_measurement_instrument) %>%
rename(instrument_name = pintens_name_measurement_instrument,
pre_m = pintens_pre_m,
post_m = pintens_post_m,
fu1_m = pintens_fu1_m,
fu2_m = pintens_fu2_m,
fu3_m = pintens_fu3_m,
instrument_present = measurements_pain_intensity)
ts_sd <- select (dat_pintens, cohort, pintens_pre_sd, pintens_post_sd, pintens_fu1_sd,
pintens_fu2_sd, pintens_fu3_sd) %>%
rename( pre_sd = pintens_pre_sd,
post_sd = pintens_post_sd,
fu1_sd = pintens_fu1_sd,
fu2_sd = pintens_fu2_sd,
fu3_sd = pintens_fu3_sd)
ts_t <- select (dat_pintens, cohort, pintens_fu1_t, pintens_fu2_t, pintens_fu3_t) %>%
rename( fu1_t = pintens_fu1_t,
fu2_t = pintens_fu2_t,
fu3_t = pintens_fu3_t)
ts_t$pre_t = -20
ts_t$post_t = 0
#change SD columns to numeric
cols.num <- c("pre_sd","post_sd", "fu1_sd", "fu2_sd", "fu3_sd")
ts_sd[cols.num] <- sapply(ts_sd[cols.num],as.numeric)
#sapply(ts_sd, class)
#convert to long dataset
# The arguments to gather():
# - data: Data object
# - key: Name of new key column (made from names of data columns)
# - value: Name of new value column
# - ...: Names of source columns that contain values
# - factor_key: Treat the new key column as a factor (instead of character vector)
ts_m_long <- pivot_longer(ts_m,
cols = c("pre_m", "post_m", "fu1_m", "fu2_m", "fu2_m", "fu3_m"),
names_to = "timepoint",
values_to = "m")
ts_sd_long <- pivot_longer(ts_sd,
cols = c("pre_sd", "post_sd", "fu1_sd", "fu2_sd", "fu2_sd", "fu3_sd"),
names_to = "timepoint",
values_to = "sd")
ts_t_long <- pivot_longer(ts_t,
cols = c("pre_t", "post_t", "fu1_t", "fu2_t", "fu3_t"),
names_to = "timepoint",
values_to = "t")
#merge long datasets
ts_sd_long$timepoint <- gsub("\\_sd*$","",ts_sd_long$timepoint)
ts_t_long$timepoint <- gsub("\\_t*$","",ts_t_long$timepoint)
ts_m_long$timepoint <- gsub("\\_m*$","",ts_m_long$timepoint)
ts_sdt_long <- merge(ts_sd_long, ts_t_long, c("cohort", "timepoint"))
ts_long <- merge(ts_sdt_long, ts_m_long, c("cohort", "timepoint"))
unique(ts_long$instrument_name)## [1] "VAS (0-10)" "VAS" "NRS (0-10)"
## [4] "NRS" "NPRS" "Likert pain intensity"
## [7] "MPI: pain severity" "NRS (0-100)" "PRI"#create max_score column
ts_long$max_scale = ts_long$instrument_name
ts_long$max_scale <- dplyr::recode(ts_long$max_scale, "VAS" = 100,
"VAS (0-100)" = 100,
"PRI" = 78,
"MPI: pain severity" = 7,
"NRS" = 10,
"NPRS" = 10,
"NRS (0-10)" = 10,
"VAS (0-10)"= 10,
"NRS (0-100)" = 100,
"Likert pain intensity" = 6,
"RDQ" = 24,
"QBPDS" = 100,
"LBPRS" = 30,
"DRI" = 1200,
"DRI (0-100)" = 100,
"MPI" = 6,
"ODI" = 100,
"ODI (0-1)" = 1,
"PDI" = 70,
"RMDQ" = 24,
"DPQ: Daily activities" = 100,
"SF-36 subscale Physical Functioning" = 100,
"NHP: PA" = 100,
"HFAQ" = 100,
"MPI: GA" = 6,
"Norfunk (0-3)" = 3,
"ADS (german scale of CES-D)" =60,
"HADS-D" =21,
"DASS" = 42,
"BDI-II" =63,
"Zung" =100,
"BDI" =63,
"depression index (DEPS)" =30,
"SCL90-D" = 52,
"HADS-A"= 21,
"VAS Anxiety (0-100)" = 100,
"SCL90-A" = 40,
"STAI" = 80,
"NHP: Emotional reactions" = 100,
"SF-36: mental health" = 100,
"DPQ: anxiety/depression" = 100,
"SCL-90: Hostility" = 24,
"PSEQ" = 60,
"DPQ: social life" = 100,
"SF-36: social functioning" = 100
)
#create reverse scoring variable
#binary variable: do higher scores indicate better functioning? 0=no; 1=yes.
ts_long$rev_score = ts_long$instrument_name
ts_long$rev_score <- recode(ts_long$rev_score, "VAS" = 0,
"VAS (0-100)" = 0,
"PRI" = 0,
"MPI: pain severity" = 0,
"NRS" = 0,
"NPRS" = 0,
"NRS (0-10)" = 0,
"VAS (0-10)"= 0,
"NRS (0-100)" = 0,
"Likert pain intensity" = 0,
"RDQ" = 0,
"QBPDS" = 0,
"LBPRS" = 0,
"DRI" = 0,
"DRI (0-100)" = 0,
"MPI" = 0,
"ODI" = 0,
"ODI (0-1)" = 0,
"PDI" = 0,
"RMDQ" = 0,
"DPQ: Daily activities" = 0,
"SF-36 subscale Physical Functioning" = 1,
"NHP: PA" = 0,
"HFAQ" = 1,
"MPI: GA" = 1,
"Norfunk (0-3)" = 1,
"ADS (german scale of CES-D)" =0,
"HADS-D" =0,
"DASS" =0,
"BDI-II" =0,
"Zung" =0,
"BDI" =0,
"depression index (DEPS)" =0,
"SCL90-D" =0,
"HADS-A"= 0,
"VAS Anxiety (0-100)" = 0,
"SCL90-A" = 0,
"STAI" = 0,
"NHP: Emotional reactions" = 0,
"SF-36: mental health" = 1,
"DPQ: anxiety/depression" = 0,
"SCL-90: Hostility" = 0,
"PSEQ" = 1,
"SF-36: social functioning" = 1,
"DPQ: social life" = 0
)
#reverse scoring
ts_long$rev_score_m <- (ts_long$max_scale + 1)-ts_long$m
#combining rev_score and normal score on contion of ts_long$rev_score
ts_long$m_final <- ts_long$m
for (row in 1:nrow(ts_long)) {
if (ts_long[row, "rev_score"] == 0 & !is.na(ts_long[row, "rev_score"])) {
# replace value by ts_long$rev_score_m
ts_long[row, "m_final"] <- ts_long[row, "rev_score_m"]
}
}
#standardize mean
ts_long$m_stand = (ts_long$m_final/ts_long$max_scale)*100
#create plot
p <- ggplot(ts_long, aes(x=t, y=m_stand, group=cohort, color=cohort, text =
paste("Author: ", author,
"<br>Year: ", year,
"<br>Cohort ID: ", cohort_id,
"<br>Instrument: ", instrument_name,
"<br>Raw mean: ", m,
"<br>Raw SD: ", sd,
"<br>Standardized score ", m_stand
))) +
labs(y= "outcome standardized (0-100)") +
geom_line(data=ts_long[!is.na(ts_long$m_stand),]) +
geom_point() +
scale_x_continuous(name = "Time (months)",
breaks = c(-20, 0, 3, 6, 12, 24, 60, 120),
labels = c("pre", "post", "3m", "6m", "12m", "24m", "60m", "120m"),
limits = c(-20, 200))
ggplotly(p, tooltip = "text") %>%
layout(
xaxis = list(
tickvals = c(-20, 0, 3, 6, 12, 24, 60, 120),
ticktext = c("pre", "post", "3m", "6m", "12m", "24m", "60m", "120m"),
ticklen = 5,
tickwidth = 2,
tickcolor = toRGB("blue"),
range=c(-22, 26)
),
yaxis = list(
range=c(0, 100))
)